VL-SAT: Visual-Linguistic Semantics Assisted Training for 3D Semantic Scene Graph Prediction in Point Cloud.
Ziqin WangBowen ChengLichen ZhaoDong XuYang TangLu ShengPublished in: CoRR (2023)
Keyphrases
- point cloud
- real world objects
- laser scanner
- semantic representations
- urban scenes
- stereo camera
- semantic information
- high level semantics
- architectural scenes
- natural language
- surface reconstruction
- structure from motion
- d scene
- dense reconstruction
- dominant plane
- multi view stereo
- point sets
- cad model
- point cloud data
- high resolution
- single image
- supervised learning
- moving objects
- video sequences
- rdf graphs
- visual features
- training samples
- linguistic expressions
- three dimensional